High-precision map construction method and device, storage medium and electronic equipment

Abstract

The invention discloses a high-precision map construction method and device, a storage medium and electronic equipment, and the method comprises the steps: determining a standard depth interval corresponding to each pixel point in a target image according to a reference depth map corresponding to the target image in a binocular image and an identification result of an object type in the target image; wherein the object category comprises a static object and a dynamic object. And according to the standard depth interval corresponding to each pixel point, filtering a depth value obtained by mapping each point cloud point to the target image to obtain a depth map corresponding to a static object in the target image, and constructing a high-precision map according to the depth map. According to the method, the depth value mapped by the dynamic object point cloud point on each pixel point is filtered according to the standard depth interval of each pixel point to obtain the depth value mapped by the static object point cloud point, so that the problem that the dynamic object point cloud point and the static object point cloud point mapped to the same pixel point are overlapped can be avoided; therefore, the accuracy of the depth value in the depth map is improved.

Description

Technical field [0001] This specification relates to the unmanned field, and more particularly to a high-precision map construction method, apparatus, storage medium, and electronic device. Background technique [0002] In the high-precision map construction process, the construction depth map is very important based on the distance between each target and the visual sensor. Among them, the higher the distance between the target and the visual sensor in the depth map, the higher the accuracy of the constructed high-precision map. [0003] In the prior art, when the depth map is constructed, each frame point cloud can be obtained by laser radar, and the target image is acquired by a single visual sensor. Then, the acquired multinorted cloud is fused, and the fused multi-frame point cloud is incident on the target image by manually calibrated outer parameter matrix, and the depth map corresponding to the target image is obtained. The depth map can represent the distance between the...

AI Technical Summary

Problems solved by technology

[0004] When the multi-frame point cloud is mapped to the target image, the point cloud of the static object and the point cloud of the dynamic object in the multi-frame point cloud are mapped to the target image, and the same pixel on the target image may have the point cloud of the static object and the point cloud of the dynamic object. The overlapping point clouds of dynamic objects lead to inaccurate depth values ​​corresponding to each pixel in the depth map obtained by multi-frame point cloud mapping

Images